Which wavelengths, power levels, and dosing parameters matter for hair growth?

Four numbers decide whether a light device on your head is doing anything at all: the wavelength, the irradiance actually reaching your scalp, the energy dose per session, and how often you repeat it. Miss any one of them and you've bought an expensive hat that warms your head on a schedule. Get all four inside the tested windows and you're running the same protocol the published trials ran.

Which light wavelengths have actual clinical evidence for stimulating hair follicles?

Most people assume any red-ish light will do, and the evidence says something much narrower. Two independent constraints happen to overlap in the same slice of the spectrum: tissue has an optical window between roughly 600 and 1100 nanometers where blood and water stop soaking up the light, and cytochrome c oxidase, the mitochondrial enzyme most people point to as the target, absorbs in the red and the near-infrared with a dead zone between them. That's why the numbers on a spec sheet either land in these bands or they don't mean anything for your follicles.

Does coherent laser light outperform LED light for the scalp?

Coherence is the property that makes laser light laser light, and it's the property that doesn't survive your scalp. Scattering in the first millimeter of tissue randomizes the phase so completely that photons arriving at a follicle several millimeters down are indistinguishable from photons that started out incoherent. Where the two formats genuinely differ is engineering, not biology, and that's the difference you should be shopping on.

How much power density at the scalp surface is needed, and how is it measured?

Irradiance is power per unit area at the target, and it's the number most often dressed up in this category. Electrical input wattage isn't optical output, the peak under one diode can run five to ten times the array average, and a hood held two centimeters off your head delivers a fraction of what the same hardware delivers in contact. Here's how the honest figure gets taken.

1. Meter at the scalp plane: Put the detector where your head sits, not at the diode face.
2. Match the detector to the wavelength: A meter that isn't sensitive at 660 or 810 nanometers reads fiction.
3. Sample across the whole array: Take readings at the crown, vertex, and hairline, not one bright spot.
4. Average rather than quote the peak: A peak reading implied as whole-scalp coverage inflates the number several times over.
5. Divide output by treated area: A helmet genuinely emitting 2000 milliwatts across 300 to 400 square centimeters averages about 5 to 7 milliwatts per square centimeter.

What total energy dose per session does the research support?

Fluence is what a protocol is really built around, and it's just irradiance multiplied by exposure time: milliwatts per square centimeter times seconds, divided by 1000, gives you joules per square centimeter. Run that arithmetic backward and you can tell whether a recommended session length makes sense or whether the manufacturer is quietly compensating for a weak device.

Can too much light make results worse rather than better?

This is where dosing stops behaving the way you expect it to. Photobiomodulation follows a biphasic dose response: the effect climbs with dose to an optimum, then falls away, and at high enough doses it inverts and suppresses the very cellular activity a moderate dose switched on. Overexposure doesn't announce itself the way a burn would, it presents as nothing happening, which is exactly what underexposure looks like too.

How often should treatment sessions run, and how long before results appear?

Frequency in the published protocols is far more consistent than dose is: every other day or three times a week, in sessions of roughly 10 to 25 minutes depending on how much irradiance the device puts out. That spacing isn't arbitrary, it keeps you off the wrong end of the dose curve and it's a schedule you might actually still be keeping in year two. Your hair cycle, not the hardware, sets everything that happens next.

1. Weeks 1 to 8: Shedding typically slows first, and a temporary bump in shedding early on is described in the literature and resolves as treatment continues.
2. Weeks 8 to 16: Finer regrowth becomes noticeable, though it's still too soon to call it.
3. Weeks 16 to 24: Terminal hair density per square centimeter becomes measurable, which is why trials set primary endpoints at 16 to 26 weeks.
4. After you stop: Gains tend to regress over the following months, since nothing about the underlying miniaturization has changed.

How much of the emitted light actually reaches the follicle bulge and bulb?

Very little of it, and every sensible protocol is built around that fact whether it says so or not. Light entering skin drops off exponentially rather than linearly, and the structures that matter aren't near the surface: the bulge holding the follicular stem cells sits between the arrector pili insertion and the sebaceous duct, and the dermal papilla of a terminal scalp follicle reaches down into the fat.

• Arrival at depth: Intensity at the bulb may be a small single-digit percentage of what hit the surface.
• Hair is the first filter: Dense dark hair absorbs and reflects much of the light before it reaches skin.
• Prep matters: Clean, dry, product-free hair with scalp contact or parted teeth beats a hood held off.
• Melanin penalty: Darker hair and skin absorb more red light, arguing for the higher end of the dose range.

Which specifications on a device datasheet are meaningful and which are marketing?

There's a short list of numbers that make a device evaluable and a much longer list that exists to look impressive. Four items let you compute delivered fluence and compare it against the literature: peak wavelength with tolerance, measured optical output at the scalp plane, treated area in square centimeters, and recommended session length. Everything else is decoration until those four are on the page.